ASSIMILATION OF REMOTE SENSING- DERIVED FLOOD EXTENT AND SOIL MOISTURE DATA INTO COUPLED HYDROLOGICAL-HYDRAULIC MODELS

Transcription

1 Patrick Matgen Giovanni Corato Laura Giustarini Renaud Hostache Stefan Schlaffer ASSIMILATION OF REMOTE SENSING- DERIVED FLOOD EXTENT AND SOIL MOISTURE DATA INTO COUPLED HYDROLOGICAL-HYDRAULIC MODELS

2 Purpose Case Study Method Research questions: How to combine remote sensing derived soil moisture and flood extent data with hydrologic-hydraulic models for improved streamflow predictions? What is the respective merit of soil moisture and flood extent data assimilation for streamflow predictions?

3 Purpose Case Study Method Hydrological and Hydraulic models represent different physical processes and take advantage of different types of observations. Meteo Forcings Hydrologic Model Hydrol DA SM Obs Different sensors are available and provide complementary information Discharge Hypothesis: assimilation of both data sets provides advantages for streamflow predictions WL, FE, Vel Hydraulic Model Hydraul DA WL, FE Obs

4 Saturation degree [-] Purpose Case Study Method Rationale: Colorso (Italy) Bibeschbach (Lux) Cordevole (Italy) Discharge [mm/h] Initiation of fast runoff is a threshold process that occurs when soil moisture rises above a critical threshold Soil moisture and water level variability are inversely correlated: potentially soil moisture and water level (or flood extent) observations are highly complementary Matgen et al., HP, 2012

5 Introduction Purpose Case Study 1 Setup of model cascade 2 Ensemble generation 3 Satellite data processing 4 Assimilation

6 Introduction Purpose Case Study The model was set up using the SuperFlex (Fenicia et al., WRR, 2010) framework Model structure can be adapted to catchment properties and satellite data characteristics Calibration performed on the basis of in situ measurements or SM satellite observations

7 Introduction Purpose Case Study LISFLOOD-FP SubGrid Designed for modelling flood flows in larger catchments. Uses DEM file as geometry. Models 1D- 2D dimensional flows or fully 2D dimensional flows. Calibration performed on the basis of in situ measurements or satellite observations

8 Introduction Purpose Case Study Precipitation 64 Particles Multiplicative Error randomly generated from lognormal distribution Ensemble of Discharge validated by the Ensemble Verification Measure of De Lannoy et al. (2006) Hydrol Mod SWI Discharge Sum of simulated Flood extent maps Hydraul Mod Flood Extents

9 Introduction Purpose Method RSD Soil M oisture ASCAT scatterometer (M ETOP satellite) Daily image acquisition Spatial resolution 2 5 km C Band RSD Flood Extent Envisat ASAR W S Period Jun Dec Spatial resolution m M ean revisit time ~ 7 days C Band AM SR-E radiometer (AQUA satellite) Spatial resolution Bi-Daily acquisition X and C band

10 Introduction Purpose Case Study Snow Masking Upscale or Downscale to Model Grid From surface to Root Zone CDF Matching Modis Snow Product 8 day aggregated AMSRE 0.25 Ascat ~ 25 km Envisat

11 Introduction Purpose Case Study The triple collocation method (Draper et al., RSE 2013) allows parameterizing the error considering a triplet of unbiased and independent datasets. Parameterization is possible if one dataset is assumed as reference For all satellite products the open loop mean was assumed as reference The following triplet was considered: Model-AMSRE-Ascat (Ascat and AMSRE error parameterization)

12 Probability of a pixel to be flooded knowing its backscatter Probability N Probability of pixels (-) Introduction Purpose Case Study UK Severn: 23 July 2007, morning UK Severn: July 2007, morning Gamma pdf (area=1) Gamma remaining pdf total (area=1) histogram part (area=1) remaining sum Gamma part of (area=1) the curve 2 pdfs sum of the 2 pdfs Probability Probability pixel value (db) pixel value (db) pixel value (db) pixel value (db) p w σ 0 p σ 0 w p(w) = p σ 0 w p w + p σ 0 d p(d) Giustarini et al., IEEE TGRS, 2013

13 Introduction Purpose Case Study Evaluation of results Flood extent obtained through photointerpretation Aerial photography (flood event on River Severn) Reliability diagram

14 Particle Filter Introduction Purpose Case Study Models States before DA Observation Spatial Weigths aggregations Weights after DA X t-1,i j W t-1,i j,0 W t,ij = Σ =0,t W,i j,0 e - t T dec Particle i 1 Particle i 2 Particle i 3 SWI t,i 1 SWI t,i 2 SWI t,i 3 SWI t,i 4 Particle 4 i Weigthed mean Particle i n-1 Particle i n SWI t,i n-1 SWI t,i n Time t- Time t Time t+

15 Introduction Purpose Case Study Assumption : binomial likelihood pdf k = # successes n = # trials MOD t,i w t, i 1,1 1,1 Simulated Water pixel Satellite observation Θ 1,1 Θ 1,2 Θ 1,3 Θ 2,1 Θ 2,2 Θ 2,3 Θ 3,1 Θ 3,2 Θ 3, t i w, t i 1, 1 w, 1, 2 t i w, t i 2, 1 w, 2, 2 t i w, t i 3, 1 w, 3, 2 w t,1 1,3 t i w, 2, 3 t i w, 3, 3 w t, i 3,3 1 3,3 w i, t j, k Simulated Dry pixel Spatial Weights aggregations w i, t j, k

16 Introduction Purpose Method Severn Catchment km 2 Cell size: Meteorological forcings: ERA-Interim In Situ Discharge Records Zambezi Catchment km 2 Cell Size: 0.7

17 Introduction Purpose Method Meteorological Forcing: ECMWF EraInterim derived products In situ discharge time series from Environment Agency for 7 gauging stations covering the period Severn at Bwedley Hydraulic Model Hydrological Model

18 Introduction Purpose Case Study Lumped E I,j P T,j Calibration period Assimilation period IR j NS = 0.74 in Asssimilation period E Q Q UT P F P FL u,j U,j Corr SWI Q I,j =0.67 (Model seems to be controllable) UR j Hydraulic model calibrated on the basis of water levels (event July 2007) P S FR Q F Q t Cell size x Distributed SR Q s Meteorological forcings (Precipitation and Temperature): ECMWF ERA Interim derived products with 6h time resolution; Precipitation spatially distributed; Evapotranspiration was computed using Hamon formula and lumped values of ERA Interim temperatures 4 Reservoirs and 10 parameters to calibrate

19 ASCAT Introduction Purpose Case Study Method

20 AMSR-E Introduction Purpose Case Study Method

21 Introduction Purpose Case Study Method Ascat AMSR-E

22 Introduction Purpose Case Study Method Flood Prob Map 23-July :27 Flood Prob Map 23-July :53

23 Introduction Purpose Case Study Method Model Sensitivity Index Variation 0 : all models in agreement Variation 1 : only half model in agreement DEM SAR Information Only Pixel showing sensitivity> 0.1 where considered for data assimilation

24 ENVISAT FE + Ascat SM Introduction Purpose Case Study Method Bewdley (upstream boundary)

25 Introduction Purpose Case Study Method Saxons Lode (intermediate gauge) +38% - 17 cm Efficiency on water elevation Absolute error on water elevation

26 Introduction Purpose Method Catchment area km 2 Meteorological Forcing: Era Interim In situ discharge time series Two Big Dams Hydraulic Model Zambezi at Tete

27 Introduction Purpose Case Study Method Rain Season from October to April Effects Cahora Bassa Dam Near constant base flow during the dry period Different response during floods

28 Introduction Purpose Case Study Method Lumped E U,j P T,j Calibration period Nov Apr 2008 Assimilation Q UT P F P URj period QOct U,j 2006 Dec FL 2009 Simulated period Jun Dec 2012 NS = 0.66 Corr SWI Q = 0.83 P S FR Q s = k s *S r0 = k s Q t Cell size 0.7 x 0.7 Hydraulic Model calibrated by NASA by SR using water levels extract Distributed from the ICESAT image acquired on Q s 13 March Reservoirs and 9 parameters to calibrate Meteorological forcings (Precipitation and Temperature): ERA Interim total precipitation and temperature 1 day time resolution; Evapotranspiration was computed using Hamon Q F

29 Introduction Purpose Method Remote sensing-derived soil moisture

30 Introduction Purpose Method AMSR-E Climatology AMSR-E Not Assimilated Correlation ASCAT Era Interim Correlation AMSR-E Era Interim Source: Dorigo et al., HESS 2010

31 ASCAT Introduction Purpose Case Study Method

32 Introduction Purpose Case Study Method ASCAT

33 Introduction Purpose Case Study Method Global Mask: Hand + Variation Index > Feb Feb Feb-2008 Hand Mask 15 m Variation Index

34 ENVISAT FE & ASCAT SM Introduction Purpose Case Study Method

35 Introduction Purpose Case Study Method The assimilation of RSD soil moisture products improves discharge predictions in specific time periods ( wet and warm ). The efficiency further depends on the satellite product, filter settings and the climatology of the catchment. The assimilation of FE outperforms SM data assimilation during storm events (when soils are saturated). RSD FE and SM provide complementary information for discharge predictions and a combination of both data sets is advantageous.

36 Introduction Purpose Case Study Method On practically all levels improvements are necessary and possible: Improve parameterisation of observing errors (e.g. autocorrelation structure, temporal and spatial variability) Further improve commensurability between models and data (necessary to fit model structures to characteristics of satellite data) Reduce existing time delay between data acquisition and higher level data dissemination (e.g. by developing fully automatic processing chains) Improvement of data assimilation schemes (e.g. feedbacks between models) Test with additional data: Sentinel-1, Cosmo Skymed, TerraSAR-X, Radarsat, ALOS, SMOS, SMAP Continue testing approach in contrasting regions across the Earth with multiple models